Analyzing the Influence of Dataset Composition for Emotion Recognition
This addresses dataset bias issues for researchers in multimodal emotion recognition, particularly in human-robot interaction, but is incremental as it analyzes existing datasets without proposing new methods.
The paper investigated how data collection methodology affects emotion recognition performance by analyzing textual dataset compositions in IEMOCAP and OMG-Emotion Behavior datasets, finding that IEMOCAP's composition negatively impacts generalization compared to OMG-Emotion Behavior.
Recognizing emotions from text in multimodal architectures has yielded promising results, surpassing video and audio modalities under certain circumstances. However, the method by which multimodal data is collected can be significant for recognizing emotional features in language. In this paper, we address the influence data collection methodology has on two multimodal emotion recognition datasets, the IEMOCAP dataset and the OMG-Emotion Behavior dataset, by analyzing textual dataset compositions and emotion recognition accuracy. Experiments with the full IEMOCAP dataset indicate that the composition negatively influences generalization performance when compared to the OMG-Emotion Behavior dataset. We conclude by discussing the impact this may have on HRI experiments.